“…P u 0 'Ho (16) It is clear that, having obtained the PMF of log L (u l1io), T is set to satisfy the Pra requirement, i.e., pia = L P(logL(u)l1io) . (17) logL(u»T Hence, the performance of the N-P fusion rule can also be obtained as follows, pi = L P(1ogL(u)l1id · (18) logL(u»T However, since log L(ul1io), is a Discrete Random Variable, its Cumulative Distribution Function (CDF) is not continuous, therefore the above equations are applicable only for some specific rates of False Alarm (would lead to a number of non randomized performance points in the (Pfa, Pd) plain.). Hence, in order to determine the N-P fusion rule and its performance for arbitrary pia, Randomization is required [19], i.e, the N-P fusion decision rule is given by, In this section we evaluate the performance of our proposed scheme through simulation.…”